[WIP] Data Intelligence Layer (DIL)

code_samples/mcp/mcp-ibexa-example-wrapper.sh

@@ -2,6 +2,7 @@
 set -e
 
 baseUrl='http://localhost' # Adapt to your test case
+mcpServer="$baseUrl/mcp/example"
 
 jwtToken=$(curl -s -X 'POST' \
   "$baseUrl/api/ibexa/v2/user/token/jwt" \
@@ -16,6 +17,6 @@ jwtToken=$(curl -s -X 'POST' \
       }' | jq -r .JWT.token)
 
 exec npx -y supergateway \
-  --streamableHttp "$baseUrl/mcp/example" \
+  --streamableHttp "$mcpServer" \
   --oauth2Bearer "$jwtToken" \
   --logLevel none

code_samples/mcp/mcp.sh

@@ -5,6 +5,7 @@ set +x
 baseUrl='http://localhost' # Adapt to your test case
 username='ibexa-example'
 password='Ibexa-3xample'
+mcpServer="$baseUrl/mcp/example"
 
 curl -s -X 'POST' \
   "$baseUrl/api/ibexa/v2/user/token/jwt" \
@@ -22,7 +23,7 @@ cat response.tmp.txt | jq
 jwtToken=$(cat response.tmp.txt | jq -r .JWT.token)
 rm response.tmp.txt
 
-curl -s -i -X 'POST' "$baseUrl/mcp/example" \
+curl -s -i -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -d '{
         "jsonrpc": "2.0",
@@ -43,15 +44,15 @@ tail -n 1 response.tmp.txt | jq
 mcpSessionId=$(cat response.tmp.txt | grep -i 'Mcp-Session-Id:' | sed 's/Mcp-Session-Id: \([0-9a-f-]*\).*/\1/i')
 rm response.tmp.txt
 
-curl -s -i -X 'POST' "$baseUrl/mcp/example" \
+curl -s -i -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -H "Mcp-Session-Id: $mcpSessionId" \
   -d '{
         "jsonrpc": "2.0",
         "method": "notifications/initialized"
       }'
 
-curl -s -X 'POST' "$baseUrl/mcp/example" \
+curl -s -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -H "Mcp-Session-Id: $mcpSessionId" \
   -d '{
@@ -60,7 +61,7 @@ curl -s -X 'POST' "$baseUrl/mcp/example" \
         "method": "tools/list"
       }' | jq
 
-curl -s -X 'POST' "$baseUrl/mcp/example" \
+curl -s -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -H "Mcp-Session-Id: $mcpSessionId" \
   -d '{
@@ -75,7 +76,7 @@ curl -s -X 'POST' "$baseUrl/mcp/example" \
         }
       }' | jq
 
-curl -s -X 'POST' "$baseUrl/mcp/example" \
+curl -s -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -H "Mcp-Session-Id: $mcpSessionId" \
   -d '{
@@ -84,7 +85,7 @@ curl -s -X 'POST' "$baseUrl/mcp/example" \
         "method": "prompts/list"
       }' | jq
 
-curl -s -X 'POST' "$baseUrl/mcp/example" \
+curl -s -X 'POST' "$mcpServer" \
   -H "Authorization: Bearer $jwtToken" \
   -H "Mcp-Session-Id: $mcpSessionId" \
   -d '{

composer.json

@@ -9,10 +9,15 @@
         }
     },
     "repositories": [
+        {
+            "type": "vcs",
+            "url": "https://github.com/ibexa/data-intelligence-layer"
+        },
         {
             "type": "composer",
             "url": "https://updates.ibexa.co"
         }
+
     ],
     "require": {
         "php": "^8.3"
@@ -99,7 +104,8 @@
         "ibexa/ckeditor-premium": "~6.0.x-dev",
         "ibexa/measurement": "~6.0.x-dev",
         "ibexa/connector-actito": "~6.0.x-dev",
-        "ibexa/fastly": "~6.0.x-dev"
+        "ibexa/fastly": "~6.0.x-dev",
+        "ibexa/data-intelligence-layer": "~6.0.x-dev"
     },
     "scripts": {
         "fix-cs": [

docs/ai/mcp/mcp_config.md

@@ -64,11 +64,27 @@
 [[= include_code('code_samples/mcp/mcp.matrix.yaml', 31, 35) =]]
 ```
 
+Servers are automatically registered as services with an ID following the pattern `ibexa.mcp.server.<repository_identifier>.<server_identifier>`.
+You can list all defined servers by running the following command:
+
+```bash
+ddev php bin/console debug:container ibexa.mcp.server
+```
+
 Routes are built automatically from MCP server `path` configs.
 Those routes are identified as `ibexa.mcp.<server_identifier>`.
 You can list them by running the following command:
 
-`php bin/console debug:router --siteaccess=<within_scope_siteaccess> ibexa.mcp`
+```bash
+php bin/console debug:router --siteaccess=<within_scope_siteaccess> ibexa.mcp`
+```
+
+### Built-in MCP servers
+
+Some MCP servers are already defined on default installation.
+
+- `data_intelligence_layer` (`/mcp/data-intelligence`) - The [Data Intelligence Layer](data_intelligence_layer.md) MCP server is bundled with tool sets from `Ibexa\Bundle\DataIntelligenceLayer\Mcp\` namespace and expose metrics about content to AI agent TODO: like the Orchestration chat one.
+  By default, it's enabled but not assigned to any SiteAccess scope. TODO: confirm; maybe it will be assigned to admin_group?
 
 ### MCP server options
 
@@ -128,6 +144,15 @@
     - `list_non_translated_content_ids` - lists IDs of content which have missing translations for a given language code.
 - `Ibexa\Mcp\Tool\SeoTools`
     - `get_non_seo_content_ids` - returns IDs of content items that are missing SEO optimization (no meta title tag). Useful for identifying content that needs SEO attention.
+- `Ibexa\Bundle\DataIntelligenceLayer\Mcp\MetricDiscoveryTools`
+    - `dil_discover_metrics` - TODO
+- `Ibexa\Bundle\DataIntelligenceLayer\Mcp\MetricReaderTools`
+    - `dil_get_metric` - TODO
+    - `dil_list_entity_metrics` - TODO
+    - `dil_query_metrics` - TODO
+    - `dil_count_metrics` - TODO
+    - `dil_get_translation_backlog` - TODO
+    - `dil_get_translation_subtree_coverage` - TODO
 
 ``` yaml hl_lines="5-7"
 [[= include_code('code_samples/mcp/mcp.matrix.yaml', 4, 7) =]]

docs/ai/mcp/mcp_guide.md

@@ -21,7 +21,7 @@
 MCP servers make the discovery of available capabilities much easier.
 They help AI agents translate natural language prompts into concrete actions on the system.
 
-![MCP communication diagram showing AI agent client connecting to MCP Server within Ibexa DXP.](img/mcp-com-diagram.png)
+![MCP communication diagram showing AI agent client connecting to MCP server within Ibexa DXP.](img/mcp-com-diagram.png)
 
 An MCP server allows the agent to discover available tools, inspect their parameters, learn how to use them, and select the correct action.
 
@@ -40,3 +40,9 @@
 
 MCP servers are defined specifically for each [repository](repository_configuration.md) and assigned to individual [SiteAccesses](siteaccess.md) scopes.
 This way you can build flexible configurations that match different contexts.
+
+## Built-in MCP servers
+
+### DIL MCP server
+
+The [Data Intelligence Layer](data_intelligence_layer.md) MCP server is an already set up server that expose tools to access metrics about content so an AI agent can work on content quality or translation coverage.

docs/ai/mcp/mcp_usage.md

@@ -101,6 +101,12 @@ In a new `config/packages/mcp.yaml` file, define a new MCP server for the `defau
 
 Adapt the `allowed_hosts` to your case, for example, if you want to use the DDEV `.ddev.site` domain instead of its `127.0.0.1` address equivalent.
 
+The server is automatically registered as a service with the ID `ibexa.mcp.server.default.example`:
+
+```bash
+php bin/console debug:container ibexa.mcp.server.default.example
+```
+
 An `ibexa.mcp.example` route is now available:
 
 ```bash
@@ -163,16 +169,16 @@ To test the `example` MCP server, a sequence of `curl` commands is used to simul
 
 `jq`, `grep`, and `sed` are also used to parse or display outputs.
 
-First, use the shell script to set the [[= product_name =]]'s base URL and user credentials as variables for easier reuse:
+First, use the shell script to set the [[= product_name =]]'s base URL, user credentials, and MCP server URL as variables for easier reuse:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 5, 7) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 5, 8) =]]
 ```
 
 Before you can communicate with the MCP server, you must first request a JWT token through the REST API:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 9, 23) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 10, 24) =]]
 ```
 
 ``` json
@@ -182,7 +188,7 @@ Before you can communicate with the MCP server, you must first request a JWT tok
 Then, perform [initialization](https://modelcontextprotocol.io/specification/2025-11-25/basic/lifecycle#initialization) to get an MCP session ID:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 21, 44) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 22, 45) =]]
 ```
 
 ``` http
@@ -196,7 +202,7 @@ Then, perform [initialization](https://modelcontextprotocol.io/specification/202
 Validate the initialization:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 46, 52) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 47, 53) =]]
 ```
 
 ``` http
@@ -206,7 +212,7 @@ Validate the initialization:
 Get the [list of tools](https://modelcontextprotocol.io/specification/2025-11-25/server/tools#listing-tools):
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 54, 61) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 55, 62) =]]
 ```
 
 ``` json
@@ -216,7 +222,7 @@ Get the [list of tools](https://modelcontextprotocol.io/specification/2025-11-25
 [Call](https://modelcontextprotocol.io/specification/2025-11-25/server/tools#calling-tools) the `greet` tool:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 63, 76) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 64, 77) =]]
 ```
 
 ``` json
@@ -226,7 +232,7 @@ Get the [list of tools](https://modelcontextprotocol.io/specification/2025-11-25
 Get the [list of prompts](https://modelcontextprotocol.io/specification/2025-11-25/server/prompts#listing-prompts):
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 78, 85) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 79, 86) =]]
 ```
 
 ``` json
@@ -236,7 +242,7 @@ Get the [list of prompts](https://modelcontextprotocol.io/specification/2025-11-
 [Get the prompt](https://modelcontextprotocol.io/specification/2025-11-25/server/prompts#getting-a-prompt) of the `greet` method:
 
 ``` bash
-[[= include_code('code_samples/mcp/mcp.sh', 87, 100) =]]
+[[= include_code('code_samples/mcp/mcp.sh', 88, 101) =]]
 ```
 
 ``` json

docs/content_management/data_intelligence_layer/data_intelligence_layer.md

@@ -0,0 +1,25 @@
+---
+description: TODO.
+month_change: true
+---
+
+# Data Intelligence Layer
+
+The Data Intelligence Layer (DIL) provides content metrics.
+
+[[= cards([
+"content_management/data_intelligence_layer/dil_guide",
+"content_management/data_intelligence_layer/dil_config",
+]) =]]
+
+TODO:
+
+- Product guide
+- Install & Config
+- User doc?
+- API:
+    - PHP API
+    - REST API
+    - GraphQL API
+    - MCP tools
+- Extend

docs/content_management/data_intelligence_layer/dil_config.md

@@ -0,0 +1,35 @@
+---
+description: TODO.
+month_change: true
+---
+
+# Data Intelligence Layer configuration
+
+## Background tasks
+
+Metric data are computed as a [background task using Ibexa Messenger](background_tasks.md).
+
+- `ibexa.data_intelligence_layer.freshness.default_interval_days`: Number of days between computation of the metrics - default is 90 days
+- `ibexa.data_intelligence_layer.freshness.content_type_intervals`: Custom number of days per content type - default is empty
+
+```yaml
+parameters:
+    ibexa.data_intelligence_layer.freshness.default_interval_days: 30 # Increase frequency to every 30 days for all content types
+    ibexa.data_intelligence_layer.freshness.content_type_intervals: # Map content type identifier to custom interval in days
+        article: 15 # Compute metrics for articles every 15 days
+```
+
+## MCP server configuration
+
+The MCP server `data_intelligence_layer` is already configured for the path `/mcp/data-intelligence`.
+It's enabled by default but not assigned to any SiteAccess.
+Assign it to the `admin_group`:
+
+```yaml
+ibexa:
+    system:
+        admin_group:
+            mcp:
+                servers:
+                    - data_intelligence_layer
+```

docs/content_management/data_intelligence_layer/dil_guide.md

@@ -0,0 +1,11 @@
+---
+description: TODO.
+month_change: true
+---
+
+# Data Intelligence Layer product guide
+
+The Data Intelligence Layer (DIL) systematically collects, aggregates, and exposes editorial and content-centric metrics.
+Editors and AI agents can use those metrics to identify possible content improvements.
+
+TODO: List available metrics and their purpose.

docs/product_guides/product_guides.md

@@ -17,6 +17,7 @@ Discover the primary ones with the help of product guides. Condensed content all
     "content_management/pages/page_builder_guide",
     "content_management/forms/form_builder_guide",
     "content_management/collaborative_editing/collaborative_editing_guide",
+    "content_management/data_intelligence_layer/dil_guide",
     "customer_management/customer_portal",
     "personalization/personalization_guide",
     "product_catalog/product_catalog_guide",

mkdocs.yml

@@ -290,6 +290,10 @@ nav:
                 - Time field type: content_management/field_types/field_type_reference/timefield.md
                 - URL field type: content_management/field_types/field_type_reference/urlfield.md
                 - User field type: content_management/field_types/field_type_reference/userfield.md
+        - Data Intelligence Layer:
+            - Data Intelligence Layer: content_management/data_intelligence_layer/data_intelligence_layer.md
+            - DIL product guide: content_management/data_intelligence_layer/dil_guide.md
+            - DIL configuration: content_management/data_intelligence_layer/dil_config.md
         - Collaborative editing:
             - Collaborative editing: content_management/collaborative_editing/collaborative_editing.md
             - Collaborative editing product guide: content_management/collaborative_editing/collaborative_editing_guide.md